1. Field
This disclosure relates generally to channel estimation and, more specifically, to techniques for improving channel estimation and tracking in a wireless communication system.
2. Related Art
As is well known, a wireless channel provides an arbitrary time dispersion, attenuation, and phase shift in a transmitted signal. While the implementation of orthogonal frequency division multiplexing (OFDM) with a cyclic prefix in a wireless communication system mitigates the effect of time dispersion caused by a wireless channel, in order to apply linear modulation schemes it is also usually necessary to remove amplitude and phase shift caused by the wireless channel. Channel estimation is typically implemented in a wireless communication system to provide an estimate (from available pilot information) of an amplitude and phase shift caused by a wireless channel. Equalization may then be employed in the wireless communication system to remove the effect of the wireless channel and facilitate subsequent symbol demodulation. Channel tracking is usually employed to periodically update an initial channel estimation. For example, channel tracking may be employed to facilitate periodic frequency-domain and time-domain channel correlation and periodic updating of channel signal-to-noise ratio (SNR), channel delay spread, and channel Doppler effect.
In a known communication system, a channel coefficient is estimated based on a training signal (pilot) and data is decoded using an estimated channel coefficient. In this system, data is subtracted from a received signal (to reduce interference with the training signal) and the channel estimation is updated based on y(n)−x^ (n), which can be repeated for a number of iterations (where y(n) corresponds to the received signal and x^(n) corresponds to a reconstructed data signal). In another known communication system, a transmission channel state is obtained from a difference between a received signal and a hard quantized value of a detected signal, which is obtained through an equalizer based on an interpolated channel estimate from scattered reference signals (pilot signals). Unfortunately, in the known communication systems, errors may be propagated as a reconstructed data signal, which is subtracted from a received signal to provide a channel estimation, may be inaccurate.